Reviews do not transfer; buyers sometimes do

The reviews are not yours. Neither is the count, the ranking, the completion history, the response-time badge or the position you spent eight months climbing to, and none of it moves with you when the account closes or you decide to leave.

This is not a complaint about platforms, which are doing exactly what they said they would. It is an accounting problem, and the useful version of it is knowing which of your assets are actually yours before you need them to be.

Sort the assets

Three piles, and the boundaries are sharper than they feel.

Theirs. Reviews, star average, ranking position, badges, completion count, tenure, and the search traffic the listing receives. All of it is data held in their system, tied to an account they control, and none of it is exportable in any form another platform would accept.

Yours, but only if you built it. Your working name, if it is consistent and used across more than one place. Your public worked example, if it lives somewhere you control. Your rate card. Your templates and rubric. Your records.

Genuinely portable. Two things, and only two. Your speed, which is the accumulated skill that took you from ninety minutes a job to thirty. And a small number of buyers who would follow a name rather than a listing.

The uncomfortable observation is that the pile doing all the persuading is the first one.

Why this is not hypothetical

Accounts in this sector close, and they close for reasons that are frequently not misconduct: a payments provider withdraws, a policy changes, a verification review goes wrong, a category is retired. Appeals are slow everywhere and the outcome is out of your hands, which is the entire argument in the piece on account risk.

Platforms also change their economics with notice, and a repricing you cannot absorb is a move you did not choose but still have to make. In the EU that notice has a floor: Regulation (EU) 2019/1150 requires platforms to notify business users of changes to their terms at least 15 days before the changes apply, which is enough time to read them and not enough to rebuild elsewhere.

So the question is not whether you will ever restart from zero. It is how much of the eight months you would take with you.

The modest things that carry across

None of these are large, and the total effort is a few hours spread over a first year.

One consistent working name, everywhere. This is the single highest-value item and it costs nothing but discipline. A name that differs by a character between two places is two names, and a buyer who remembers you cannot find you. Cross-platform consistency is worth getting right at the point you choose the name rather than later.

A public worked example you host. The generic sample, on material nobody owns, that you can point at from any listing on any platform. It is the only piece of evidence you can rebuild a cold listing around, and it is the cheapest thing to make in a quiet week.

Your own records of what you did. Job counts, formats, turnaround, take-home. You cannot present these as reviews, and they let you set a rate card on day one somewhere new rather than guessing again. This is one of the arguments in the case for keeping records that only pays off years later.

A second, established listing. Not a half-built one, which is worse than none. A real second presence takes months to become useful and cannot be created in the week you need it, which is why the timing of a second platform is a decision to make while things are going well.

A permitted route back to repeat buyers. Modest and heavily constrained. Direct contact is prohibited in most terms and worth reading carefully before doing anything at all, and where a legitimate route exists it is worth building slowly. A handful of buyers who follow a name is a realistic outcome. A transferable audience is not.

What is not worth doing

Do not screenshot your reviews. They persuade nobody on a new listing, because a screenshot of praise is indistinguishable from a screenshot of invented praise, and presenting them tends to read as a person explaining why they had to leave.

Do not spread yourself across four platforms to hedge. Four thin listings earn less than one established one and quadruple the admin, and the dilution costs you the ranking that was the point.

Do not build a public personality as insurance. An audience is a second job with its own skills and a slow payoff, and choosing it as a hedge rather than because you want it is how people end up doing two things badly.

The part that genuinely transfers

Your judgement and your minutes per job.

Those are the assets that produced the reviews in the first place, and they are the reason a restart takes weeks rather than the eight months the first one took. Someone rebuilding a listing with a working rubric, a tested rate card and thirty minutes a job is a different proposition from someone starting, whatever the count says.

A defensible method is a large part of that, and precision about how you assess is the piece that survives every move; the documented measurement conventions are where to check yours against something outside your own habits. The generic layer will not transfer because it does not need to - anything a scoring tool produces free and instantly was never the part buyers were paying you for. And what makes a specific person worth commissioning rather than a listing worth clicking is described plainly from the buyer's side, which is the closest thing to a portable asset this trade has.

If you are choosing where to build the deep listing rather than the hedges, the thing to compare is what the platform actually gives an earner in return for its cut, and Rate Cock's judges page sets out its side of that.